Multithreaded Convolution for Edge Detection

This project performs edge detection on images using a multithreaded convolution algorithm.

Prerequisites

• OpenCV 4
• C++11 compliant compiler
• pthread library

Ensure that all prerequisites are installed before proceeding.

Compilation

Using Makefile

From the project's root directory, open a terminal and run:

make

Manual Compilation

Alternatively, compile manually with:

g++ -std=c++11 -O2 -Wall -o edge_detection src/main.cpp `pkg-config --cflags --libs opencv4` -lpthread

Running the Program

After a successful compilation, run the program using:

./edge_detection <input_image> <num_threads>

For example, to process different images with 4 threads:

./edge_detection images/1MP.jpg 4
./edge_detection images/2MP.jpg 4
./edge_detection images/4MP.jpg 4
./edge_detection images/8MP.jpg 4
./edge_detection images/16MP.jpg 4
./edge_detection images/32MP.jpg 4

The command takes the specified input image (e.g., images/1MP.jpg), processes it using the designated number of threads (e.g., 4), and saves the output image as output_1MP.jpg in the current directory.